See: https://open.substack.com/pub/garymarcus/p/three-reasons-to-think-that-the-claude?r=8tdk6&utm_medium=ios

In an April 9, 2026 Substack post, AI skeptic Gary Marcus argues that Anthropic's announcement of its Claude "Mythos" model was significantly overhyped, offering three key reasons for skepticism: the Firefox exploitation demo had sandboxing disabled, making it a proof-of-concept rather than a real-world threat; open-weight models can already replicate much of what Mythos demonstrated; and benchmark data shows Mythos is only incrementally better than prior models, tracking closely with existing ECI trends rather than representing a breakthrough. Marcus concludes that while the announcement underscores the need for better regulatory and technical safeguards, it does not represent the immediate, off-the-charts danger that media coverage implied.

No, Mythos is not that good; its PR is that good. 🙄
Anthropic's Claude Mythos isn't a sentient super-hacker, it's a sales pitch — claims of 'thousands' of severe zero-days rely on just 198 man…
Some sober thinking about Mythos (full version with links at my newsletter): 1It’s probably not as bad as they say, as AI and cybersecurity …
People in Washington get played, yet again We really should worry about cybersecurity - a lot – but Mythos is not the model these guys think…
Claude Mythos is Delusional

Source: model-releases

Related